ImageRecommendedGemini 3.1 Flash Image Preview
Nano banana 2 — extreme ratio Gemini image generation with optional Google search grounding.
Free text · required · ≤ 1000 chars
Ten ratios from 1:1 to 21:9
Nano banana renders at 1K only
1–4 images per request
Up to 14 · public HTTPS URLs only
Try one of these prompts
Real-world workflows and production use cases you can build and ship with this model.

Edit existing e-commerce shots without a reshoot. Pass the original URL in image_urls, describe the change in plain English, and Nano banana returns an updated render that keeps the product geometry intact while the surrounding scene gets reworked. Run a hundred SKUs through the same prompt template and the catalog stays visually consistent — same lighting language, same background style, every variant on brand.
Edit a photo
Spin four hero variants from a single prompt with n=4 and pick the winner before stand-up ends. The default Nano banana id is tuned for speed and cheap drafts; switch to `gemini-2.5-flash-image-preview-official` when an investor deck or paid placement needs the higher-fidelity render. Same endpoint, same JSON body — only the model id changes between drafts and finals.

Pass up to 14 reference images and describe how they should combine. Nano banana handles subject + scene + style references in a single call — drop a product photo plus a moodboard frame plus a brand palette swatch and the model fuses them into a new on-brand shot. No mask file, no separate compositing step. Public HTTPS URLs only — upload to your own R2 / S3 / OSS first.
Credit-based — 1 credit = $0.001 USD. Pay only for completed generations.
| Category | Unit | Price |
|---|---|---|
| Default | ||
| 1K | 1 image | $0.015 15 credits |
| Official Google | ||
| 1K | 1 image | $0.036 36 credits |
Send model `gemini-2.5-flash-image-preview` for the default Nano banana, or `gemini-2.5-flash-image-preview-official` for Google's official endpoint. Either id hits the same `/api/v1/images/generations` route — switch by changing one string in the request body. No second SDK, no second key, no separate billing surface to wire up. Nano banana behaves like any other reApi model from your code's point of view.
Pricing is in reAPI credits where 1 credit equals $0.001 USD. The estimator on this page costs your exact configuration before submit, so the Nano banana charge in the dashboard matches what the playground showed. No metering surprises, no minimum commitment, no monthly platform fee — only what your app actually generates with Nano banana, billed per image.
Submit a Nano banana task, get a task_id back instantly, poll until the result lands. Serverless functions, queue workers, edge runtimes — wherever you run, the async pattern fits. Failed and moderated requests are not charged, so a flaky prompt never costs you credits. The same lifecycle covers every reApi model — Nano banana plugs into your existing task plumbing.
Sign up at reApi and grab a key from the dashboard. Free signup credits cover the first hundred-plus Nano banana generations, enough to ship the first feature and watch the bill stay at zero while you iterate.
OpenPOST to /api/v1/images/generations with model: gemini-2.5-flash-image-preview and a prompt. Add size, n, or image_urls as you need — the gateway validates every field before charging. Nano banana returns a task_id immediately so the request never blocks your handler.
OpenGET /api/v1/tasks/:id until status is completed. The Nano banana response carries image URLs on our CDN, stable for 24 hours — copy them to your own storage if you need them longer than that.
OpenCommon questions about this model.
Explore more models in the same category.
ImageRecommendedNano banana 2 — extreme ratio Gemini image generation with optional Google search grounding.
ImageRecommendedNano banana Pro — high-fidelity Gemini image generation up to 4K with up to 14 reference images.
ImageImagen 4.0 — Google's text-to-image at a flat per-image rate across five aspect ratios.
ImageAlibaba
Qwen Image 2 — Alibaba's unified text-to-image and editing model with strong in-image text rendering, on one OpenAI-compatible endpoint.
Try it in the playground or grab an API key to integrate now.